Many applications exist that require the linear position of an object to be estimated. In some applications, very precise estimates are needed. One way to estimate object position involves the use of a magnet and a magnetic field sensor. The magnet may be attached to the movable object and the sensor may be held in a fixed position proximate to the magnet. The position of the object relative to the fixed sensor may then be estimated based on the magnetic field readings of the sensor. A problem that arises with this approach is that various factors can compromise position estimate accuracy. For example, manufacturing tolerances (e.g., variability in the air gap between the sensor and the magnet, etc.) can result in errors in the position estimates generated in a system. Similarly, component tolerances (e.g., variability in magnet strength, etc.) can result in position estimate errors. Environmental factors, such as temperature variability and stress effects, may also negatively affect position estimation accuracy in such systems. Position estimation techniques are needed that are more tolerant to some or all of the above described sources of error.